The International Journal Of Engineering And Science (IJES)

|| Volume || 5 || Issue || 10 || Pages || PP 01-7 || 2016 ||

ISSN (e): 2319 1813 ISSN (p): 2319 1805

Application of Reliability Analysis for Predicting Failures in

Cement Industry
Alexander N. Okpala and Sini R. Yelebe
Department of Mechanical Engineering, Niger Delta University, Wilberforce Island Bayelsa State, Nigeria.
-------------------------------------------------------- ABSTRACT------------------------------------------------------------This research entails the use of reliability analysis for predicting failures of machines used in the cement
industries and was done by evaluating machine down times data. This research work was necessitated by the
need to accurately predict failures of the machines used in the cement industries and come up with an effective
planning, for preventive maintenance schedule and reducing down times through developed mathematical
model for the machines. The failure frequency variation with time was determined and a regression analysis
using least squares methods. Correlation was done to ascertain the suitability of linear regression of the data
and also to determine that, the independent variable is a good predictor of the dependent variable. The
reliability model of the machines was achieved by applying the down times and the regression analysis result of
the machines studied for a period of six years to the Weibull model. Two critical components of the machines
were identified; contributing a total of 55 % of the down time. It was concluded that the critical components
indicate the trend of failure of the machines. Therefore, reducing the failure rate of these components will
increase the useful life of the machines and the obtained failure ratemodel, could be used as an important tool
for predicting future failures and hence, effectively planning against such failures.
Keywords: Reliability, Failure rate model, Prediction, Weibull, Critical components, Down-time.
--------------------------------------------------------------------------------------------------------------------------------------Date of Submission: 17 September 2016
Date of Accepted: 08 October 2016
--------------------------------------------------------------------------------------------------------------------------------------

I.

INTRODUCTION

Reliability engineering is a tool used to define the probability that a unit component or system will perform its
intended function uninterruptedly, under a specified operational working condition over a given period of time.
According to Shakuntla et al reliability analysis helps us toobtain the necessary information about the control of
various parameters. According to (Barringer and Barringer, 1997), reliability plays an important role in the
selection of equipment for lowest long term cost of ownership.The study ofthe reliability of various units in
production industry will yield a high level of machine performance with consistency, andit is important to study
the variation of reliability with respect to time,(Deepika and Kuldeep, 2009).With a growing competitive
market, there is the need to cut down cost of production, (Khan and Darrab, 2010), improve productivity and
delivery performance of manufacturing systems which are important to the industries, (Hani, 2012). (Okpala and
Kotingo, 2007)as well describe the reliability evaluation of a product or processto include a number of different
analysis, depending on the phase of the product life cycle. (Damilare and Olasunkanmi, 2010) described the
failure of an item entering a given age and time interval as the conditional probability, while the failure
increases with time. Building a reputation on reliability takes a longtime, and only a short while for unreliability
(Okpala and Kotingo, 2007). According to (Igor, 2004), the price for unreliability is very high, and reliability is
the cure. According to(OConnor and Kleyner, 2012), unreliability is so high a prize to be paid by any industry,
and the survival of any industry is reliability.
The reliability of mechanical components is usually estimation from test or experience by mean time to failure
(MTTF) for each machine. Several works exist showing the relationship between machine reliability and
production level, and that productivity is tied to machine system reliability. Machine reliability is assessed by
three phases of failure, the early life phase, useful life phase and the wear out phase. The application of
reliability engineering analysis is a viable tool to machine satisfactionand the constant practice of machine
reliability will always sustain a company with its products. Hence, there is need for machine reliability, which is
the live wire for the sustenance of every business. This research dwells on identifyingthe three stages of failure
and development of model for predicting the likely failure times of the various machine components under
study.

www.theijes.com

The IJES

Page 1

Application Of Reliability Analysis For Predicting Failures In Cement Industry

II.

METHODOLOGY

Data Collection
Reliability data was collected from the maintenance report books of work done procedures in a cement
production industry in south-south zone of Nigeria. The data covers a period of six years from 2008 to 2014,
which consist of themachine down-time and the number of failures for two identical machines code named BE
machine 01 and BE machine 02. The BE machine 01 and 02 hasfourteen components namely;segment wheel
drive (SWD), chain, shackle, shaft, twin chute, bearing, heavy duty gear box (HDGB), electric motor, wheel,
flange set, counter weight, turbo and hub.Thesecomponents are arranged in series, thatis, the failure of one
means the failure of the entire machine as production would stop.
Procedure
The data obtained from the maintenance report books of work done procedures of the cement production
industrywere used for evaluating the total down time and the failure rates of the machines over six years. Also,
through the Weibull model the data were subjected to life testing with aresultant modelling of the failure rate for
the machines.
Total Down-Time
The total down-time for the machines over the study period of six yearswas evaluated. This was determined by
summing up the monthly downtimes of all eighteen components of both machines over the study period of six
years(2008 2014)and presented in tables 1 and 2.
Table 1: Total Down-Time of machine BE 01
Component
SWD
Chain
Shackle
Shaft
Twin Chute
Bearing
HDGB
Electric Motor
Wheel
Flange Set
Counter Weight
Turbo
Bucket
Hub

DT2009
(min)
0
134
164
0
0
47
0
296
0
345
0
163
611
0

DT2010
(min)
0
0
385
0
0
102
0
311
0
509
0
0
648
0

DT2011
(min)
0
107
291
0
0
0
0
205
145
502
0
0
696
0

DT2012
(min)
0
133
398
0
0
105
0
368
188
403
0
0
647
0

DT2013
(min)
0
80
175
0
0
0
0
188
0
198
0
0
351
0

DT2014
(min)
0
290
96
0
0
230
0
0
79
255
0
268
659
0
TOTAL

TDT
(min)
0
744
1509
0
0
484
0
1368
412
2212
0
431
3612
0
10772

Table 2: Total Down-Time of machine BE 02

Component
SWD
Chain
Shackle
Shaft
Twin Chute
Bearing
HDGB
Electric Motor
Wheel
Flange Set
Counter Weight
Turbo
Bucket
Hub

III.

DT2009
(min)
0
217
205
73
0
47
0
53
0
444
0
85
488
0

DT2010
(min)
0
177
140
0
0
0
0
174
28
420
0
231
563
0

DT2011
(min)
0
136
168
0
0
212
0
247
0
269
0
0
510
0

DT2012
(min)
201
160
317
0
0
209
0
234
124
321
0
205
520
0

DT2013
(min)
0
179
262
186
0
83
0
140
16
406
0
199
434
0

DT2014
(min)
0
345
234
0
0
257
0
0
0
297
0
17
1079
0
TOTAL

TDT
(min)
201
1214
1326
259
0
808
0
848
168
2157
0
737
3594
0
11312

FAILURE FREQUENCY OVER TIME

The data obtained from the maintenance report books of work done procedures in a cement production
industrywere used to determine the failure frequency variation with time. This was done by dividing the
machine cumulative number of failures by the cumulative time of use of machine in hours to obtain tables 3 and
4 which were used in plotting the failure rate graphs for the machines in figures 1 and 2.

www.theijes.com

The IJES

Page 2

Application Of Reliability Analysis For Predicting Failures In Cement Industry

Table 3: Failure frequency over time for BE 01
Year
Cumulative time of use (hours)
BE 01
Cum. failure freq.
Failure Rate (X10-3)
Rubber Coupling
Cum. failure freq.
Failure Rate (X10-3)
Bag Cleaner
Cum. failure freq.
Failure Rate (X10-3)

2009
8760
88

2010
17520
179

2011
26280
274

2012
35064
399

2013
43824
466

2014
52584
559

10.05
16

10.22
32

10.43
48

11.38
68

10.63
79

10.63
94

1.83
9

1.83
19

1.83
29

1.94
42

1.80
55

1.79
64

1.03

1.08

1.10

1.20

1.26

1.22

Table 4: Failure frequency over time for BE 02

Year
Cumulative time of use (hours)
BE 02
Cum. failure freq.
Failure Rate (X10-3)
Bucket
Cum. failure freq.
Failure Rate (X10-3)
Flange Set
Cum. failure freq.
Failure Rate (X10-3)

2009
8760
133

2010
17520
251

2011
26280
379

2012
35064
598

2013
43824
793

2014
52584
923

15.83
20

14.32
42

14.42
62

17.05
94

18.10
124

17.55
144

2.28
13

2.40
25

2.36
38

2.68
60

2.82
82

2.74
94

1.48

1.43

1.46

1.71

1.87

1.79

Figure 1: Graph of Failure Rate over Time of machine BE 01 and Critical Components

Figure 2: Graph of Failure Rate over Time of machine BE 01 and Critical Components

www.theijes.com

The IJES

Page 3

Application Of Reliability Analysis For Predicting Failures In Cement Industry

IV.

MODELLING OF THE FAILURE RATE

According to (OConnor and kleyner, 2012), it is useful in engineering to determine which distribution that best
fits a set of data and to derive estimates of the distribution parameters.The Weibull model was used as the
basisfor the modelling of the failure rate. (Lyonnet, 1991) stated that the Weibull model is the most suitable
when carrying out reliability analysis for mechanical components and (OConnor and kleyner, 2012) also stated
that the Weibull probability data analysis is the most utilized technique for processing and interpretation of life
data.
According to (OConnor, 1991), the Weibull failure rate function is given by
Where:
= scale parameter,
= the shape parameter, also known as the Weibull slope.
t = variable time.
The failure density function according to
(Lyonnet, 1991), is given as
and the reliability function is given as
the Weibull distribution is:

According to (OConnor and Kleyner, 2012) the Weibull distribution is given by

which transforms to

The above equation is of a linear form

Therefore, equation 6 can thus be written as

Therefore, using the recommended estimator by(Lyonnet, 1991) for the behavior of part failure, the values of
the various parameters can be estimated. Thus, F(t) can be obtained by the given equation below.

Where n= sample times = 6, which corresponds to the number of years of study. i= the year from 1- 6
Therefore, the first year: i= 1, and n = 6.

For the second year: i= 2, and n = 6.

For the third year: i= 3, and n = 6.

For the fourth year: i= 4, and n = 6.

For the fifth year: i= 5, and n = 6.

www.theijes.com

The IJES

Page 4

Application Of Reliability Analysis For Predicting Failures In Cement Industry

For the sixth year: i= 6, and n = 6.

V.

LEAST SQUARES AND LINEAR REGRESSION

To determine the values of and for machinesBE 01, the methods of least squares is used with the established
linear relation Y = x + ln. The least square estimation applies to the n observation pairs of (xiyi), where i= 1,
2, 3,, n which leads to the estimates.
A regression analysis is done and the formulation of tables 5 and 6 for the machines with the transformed linear
function.
Where xi = ln F(t), and yi = ln ti
With the corresponding values forF(t), ti and yi regression analysis table was evaluated for BE 01 machine.
Table 5: Regression Analysis for machine BE 01
F(t)
0.08
0.25
0.42
0.58
0.75
0.92

ti
8730
17439
26163
34914
43636
52355

F(t)
0.08
0.25
0.42
0.58
0.75
0.92

ti
8733
17464
26198
34944
43673
52395

xi
9.08
9.77
10.17
10.46
10.69
10.87
xi=61.04

yi
-2.53
-1.39
-0.87
-0.55
-0.29
-0.08
yi=-5.71

xiyi
-22.97
-13.58
-8.85
-5.75
-3.1
-0.87
xiyi=-55.12

X2
82.45
95.45
103.43
109.41
114.28
118.16
X2=623.18

Y2
6.4
1.93
0.76
0.3
0.08
0.0064
Y2=9.48

Table 6: Regression Analysis for machine BE 02

xi
9.08
9.77
10.17
10.46
10.69
10.87
xi=61.04

yi
-2.53
-1.39
-0.87
-0.55
-0.29
-0.08
yi=-5.71

xiyi
-22.97
-13.58
-8.85
-5.75
-3.1
-0.87
xiyi=-55.12

X2
82.45
95.45
103.43
109.41
114.28
118.16
X2=623.18

Y2
6.4
1.93
0.76
0.3
0.08
0.0064
Y2=9.48

Correlation is determined to ascertain the suitability of linear regression for this data.
According to (Stevenson 1991).

From the above value, the independent variable is a good predictor of the dependent variable, therefore
regression is very suitable.
The slope

Y intercept, ln is given by

www.theijes.com

The IJES

Page 5

Application Of Reliability Analysis For Predicting Failures In Cement Industry

Then for BE 01

Where

is gamma function obtained from standard tables to be = 0.9086

For Machine BE 02,

Table 5 and Table 6 for BE 01 and BE 02 respectively are almost the same apart from the second column that
differs. The other six remaining columns for both machines are the same. Hence, the failure rate model and the
mean time to failure for machines BE 01 and BE 02 are the same.
Reliability Model for the Machine
The reliability model is attained by recalling the modelled reliability Weibull distribution equation and
substituting for the values of and for BE 01 machine intoequation 2.Therefore, the needed reliability model
for the BE machines is given as

VI.

DISCUSSION OF RESULTS

Total Down-Time of Machine

Table 1 and Figure1 presented results of the total down-time of machine BE01 over the six years period of
study, with a total down-time of 10,772minutes. The two components with the highest percentage of downtimes are the bucket and flange set, with each contributing 34% (3,612 minutes) and electric drive drum 21%
(2,212 minutes) respectively making a total contribution of 55% to the total down-time. Observation of Table 2
and Figure2 also show a similar trend but this time contributing 31% to the total down-time of machine BE 02
which has a total downtime of 11,312 minutes, with bucket contributing 32% (3,594 minutes) and flange set
19% (2,157 minutes) respectively to the machine total down-time. The two components bucket and flange set
with greater inputs to the machine down-time are the critical components of the machine, and the reduction of
the down-times of these critical components would equally reduce the down-times of the machine.
Failure Rate Graph for Machine
The Bath-tub curve which is the ideal failure rate graph can be used to tell the operational state of any system.
The curve may be divided into three stages, the early life stage with a decreasing failure rate, the useful life
stage with a constant failure rate and the wear out stage with an increasing failure rate. Comparing a systems
failure rate curve with that of the Bath-tub will give the stage at which a particular system is operating. The
graph in Figure 1 shows that the machine (BE 01) and its flange set are still within its useful life stage, the
bucket appears to have erratic failure rate tending towards the wear out stage. Observation of the graph of Figure
2 shows a similar trend with the slope of the flange set being similarto that of the machine and indicating that
the failure rate of the flange set is adversely affecting the machine with a constant failure rate. The slope of the
machine shows that the machine is tending towards the wear out portion, reducing the failure rate of the flange
set will put the machine back to its useful life stage.
Modelling
The model obtained for machines BE 01 and BE 02, could serve as a very important tool in extrapolating the
likely failure rate with time of each of these machines and hence, institute planned maintenance to reduce the
down-time of the machines, save cost and increase productivity. A failure rate model obtained for any system
can be used to predict when failure

VII.

CONCLUSION AND RECOMMENDATION

This study looked at reliability engineering and shows how to predict reliability and demonstrate it with test and
field data. Model was also developed with which machine failure in the cement industry in Nigeria could be
predicted. Two components were identified as been critical for the machine operation. Thesaddle set and

www.theijes.com

The IJES

Page 6

Application Of Reliability Analysis For Predicting Failures In Cement Industry

impeller blade, bothdepicts the trend of the machines failure. Therefore, reducing the failure rate of the two
critical components of the machines will increase the useful life of the machines. It is recommended for the
manufacturer to have these components arranged in parallel rather than in series, as this will in turn appreciate
the useful life of the machines since the failure of a component will kick-start the parallel pair to take over its
function and reduce the overall down-time of the machine.
The failure rate model obtained could serve as a working tool intheforecast or prediction of further future failure
rate in the machinesand hence instituting a planned maintenance.

REFERENCES
[1]. Shakuntlaa, Sanjay, K., Lala, A. K., Bhatiaa, S.S., and Jai Singhc, [2011], Availability Analysis of Polytube Industry When Two SubSystem Are Simultaneous Fail. Bangladesh Journal of Scientific and Industrial Research. 46(4), 475-480.
[2]. Barringer, H.P., Barringer, P.E. & Associates, Inc., [1997], Life Cycle Cost & Reliability for Process Equipment. Conference &
Exhibition, Houston, Texas. PennWell Publishing.
[3]. Deepika, G. and Kuldeep, K., [2009], Reliability Analysis of Pharmaceutical Plant Using Mat lab-Tool.International Journal of
Electronics Engineering Research Volume 1 Number 2 (2009) pp. 127133. Research India Publications.
[4]. Khan, M.R.R. and Darrab, I.A., [2010], Development of analytical relation between maintenance, quality and productivity, Journal of
Quality in Maintenance Engineering, 16.
[5]. Hani Shafeek [2012], Maintenance Practices in Cement Industry. Asian Transactions on Engineering Volume 01 Issue 06
[6]. Okpala, A. N. and Kotingo, K, 2007 Application of Reliability Analysis for Predicting Failures in Glass Industry Global Journal of
Engineering Research vol. 6. No. 2
[7]. Damilare T. O. and Olasunkanmi O. A, [2010], Development of Equipment Maintenance Strategy for Critical Equipment. Pacific
Journal of Science and Technology. 11(1)
[8]. Igor B. [2004], Reliability Theory and Practice, Courier Dover Publication, Minneola, NY.
[9]. Lyonnet, P., [1991], Tools of Total Quality, An Introduction to statistical Process Control. Chapman and Hall, London.
[10]. OConnor,P. D. T. and Kleyner, A., [2012], Practical Reliability Engineering,John Wiley and Sons,Chichester, UK, 5th edition,
[11]. O'Connor,P. D. T. [1991],Practical Reliability Engineering, John Wiley & Sons, Chichester, UK, 3rd edition,
[12]. Stevenson, W.J., [1991], Production Operational Management. 6thedition

www.theijes.com

The IJES

Page 7